Notifications
Clear all
Topic starter
June 27, 2026 5:00 am
I just set up my first local LLM on a Pi cluster. I was excited to try some community "skills" to give it more functions.
But looking at the install scripts... some just curl | bash from random domains. Others ask for full system access right away. This feels like the opposite of the "open" and "security-first" ideas we talk about here.
Maybe I'm missing something. How do you vet these before running them? Is there a safe way to try them at all?